INDEX

Explanations

names related to characters or objects in a specific show or series

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

mentions of "Cart" or related terms in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ITED

-0.61

 bond

-0.57

sow

-0.55

 Banks

-0.55

 chancellor

-0.55

ited

-0.55

 bonds

-0.55

 dice

-0.53

 pinch

-0.53

 Pact

-0.51

POSITIVE LOGITS

ilage

1.50

ographer

1.35

ographers

1.32

esian

1.30

wright

1.27

oons

1.27

ridges

1.20

wheel

1.11

ography

1.11

ridge

1.09

Activations Density 0.015%

names related to characters or objects in a specific show or series

mentions of "Cart" or related terms in various contexts

No Comments

No Known Activations